Node Classification on Citeseer standard (test) (Perturbation Robustness)
75.5Accuracy (Clean)RUNG-l1
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| RUNG-l1Attack Type=Adaptive global attack2023.11 | 75.5 | 69.3 | 65.9 | 61.1 | 57.2 | 53.9 | |
| APPNPAttack Type=Adaptive global attack2023.11 | 75.3 | 65.8 | 60.7 | 60.4 | 46 | 41.2 | |
| GCNAttack Type=Adaptive global attack2023.11 | 74.8 | 66.1 | 60.9 | 53 | 47 | 41.2 | |
| GRANDAttack Type=Adaptive global attack2023.11 | 74.8 | 66.6 | 61.8 | 53.6 | 47.4 | 42.2 | |
| Jaccard-GCNAttack Type=Adaptive global attack2023.11 | 74.8 | 66.3 | 60.9 | 53.3 | 46.5 | 41.1 | |
| GARNETAttack Type=Adaptive global attack2023.11 | 74.8 | 68 | 64 | 58.2 | 53.9 | 51 | |
| EvenNetAttack Type=Adaptive global attack2023.11 | 74.6 | 66.8 | 62 | 55.9 | 51.1 | 47.4 | |
| SoftMedianAttack Type=Adaptive global attack2023.11 | 74.6 | 68 | 64.4 | 59.3 | 55.2 | 51.9 | |
| RGCNAttack Type=Adaptive global attack2023.11 | 74.4 | 66 | 60.6 | 52.5 | 46.1 | 40.2 | |
| RUNGAttack Type=Adaptive global attack2023.11 | 74.3 | 71.4 | 69.8 | 67.6 | 66.5 | 65.3 | |
| ProGNNAttack Type=Adaptive global attack2023.11 | 74.2 | 65.6 | 60.3 | 52.7 | 46.2 | 40.8 | |
| TWIRLSAttack Type=Adaptive global attack2023.11 | 74.2 | 69.2 | 66.4 | 61.6 | 58.1 | 51.8 | |
| TWIRLS-TAttack Type=Adaptive global attack2023.11 | 73.7 | 69.1 | 66.4 | 62.8 | 60 | 57.4 | |
| GATAttack Type=Adaptive global attack2023.11 | 73.4 | 65.4 | 60.4 | 52.6 | 47.2 | 41.2 | |
| GNNGuardAttack Type=Adaptive global attack2023.11 | 72.4 | 65.6 | 61.8 | 55.6 | 51 | 47.3 | |
| MLPAttack Type=Adaptive global attack2023.11 | 67.7 | 67.7 | 67.7 | 67.7 | 67.7 | 67.7 |